Abstract
The processing of stationary sounds relies on both local features and compact representations. As local information is compressed into summary statistics, abstract representations emerge. Whether the brain is endowed with distinct neural architectures overseeing such computations is unknown.
In this magnetoencephalography (MEG) study, we employed a validated protocol to localize cortical correlates of local and summary representations, exposing participants to triplets of synthetic sound textures systematically varying for either local details or summary statistics. Sounds also varied for their sound duration, specifically short (40ms) or long (478ms). Results revealed clear distinct activation patterns for local features and summary statistics changes. Such activations diverged in magnitude, spatiotemporal distribution, and hemispheric lateralization. For short sounds, a change in local features, compared to summary statistics, predominantly activated the right hemisphere. Conversely, for long sounds, a change in summary statistics elicited higher activation than a change in local features in both hemispheres.
Specifically, while the right auditory cortex was responding more to changes in local features or summary statistics depending on sound duration (short or long, respectively), the left frontal lobe was selectively engaged in processing a change in summary statistics at a long sound duration. These findings provide insights into the neural mechanisms underlying the computation of local and summary acoustic information and highlight the involvement of distinct cortical pathways and hemispheric lateralization in auditory processing at different temporal resolutions.
Significant Statement We revealed hemispheric specializations for auditory computations at high (local) and low (summary statistics) temporal resolutions. The right hemisphere was engaged for both computations, while the left hemisphere responded more to summary statistics changes. These findings highlight the multifaceted functions of the right hemisphere in capturing acoustic properties of stationary sounds and the left hemisphere’s involvement in processing abstract representations.
In this magnetoencephalography (MEG) study, we employed a validated protocol to localize cortical correlates of local and summary representations, exposing participants to triplets of synthetic sound textures systematically varying for either local details or summary statistics. Sounds also varied for their sound duration, specifically short (40ms) or long (478ms). Results revealed clear distinct activation patterns for local features and summary statistics changes. Such activations diverged in magnitude, spatiotemporal distribution, and hemispheric lateralization. For short sounds, a change in local features, compared to summary statistics, predominantly activated the right hemisphere. Conversely, for long sounds, a change in summary statistics elicited higher activation than a change in local features in both hemispheres.
Specifically, while the right auditory cortex was responding more to changes in local features or summary statistics depending on sound duration (short or long, respectively), the left frontal lobe was selectively engaged in processing a change in summary statistics at a long sound duration. These findings provide insights into the neural mechanisms underlying the computation of local and summary acoustic information and highlight the involvement of distinct cortical pathways and hemispheric lateralization in auditory processing at different temporal resolutions.
Significant Statement We revealed hemispheric specializations for auditory computations at high (local) and low (summary statistics) temporal resolutions. The right hemisphere was engaged for both computations, while the left hemisphere responded more to summary statistics changes. These findings highlight the multifaceted functions of the right hemisphere in capturing acoustic properties of stationary sounds and the left hemisphere’s involvement in processing abstract representations.
Original language | English |
---|---|
Publisher | bioRxiv |
Number of pages | 31 |
DOIs | |
Publication status | Published - 4 Aug 2023 |
Fields of Science and Technology Classification 2012
- 501 Psychology